OBSOLETE Patch-ID# 119966-01


Download this patch from My Oracle Support

Your use of the firmware, software and any other materials contained in this update is subject to My Oracle Support Terms of Use, which may be viewed at My Oracle Support.
For further information on patching best practices and resources, please see the following links:
Copyright (c) 2012, Oracle and/or its affiliates. All rights reserved.

Keywords: libm libmvec gcc us3
Synopsis: Obsoleted by: 119966-02 SunOS 5.10: Math Libraries (libm+libmvec) patch
Date: Mar/12/2010


Install Requirements: NA

Solaris Release: 10

SunOS Release: 5.10

Unbundled Product:

Unbundled Release:

Xref: This patch available for x86 as patch 119967

Topic:

Relevant Architectures: sparc

Bugs fixed with this patch:

Sun CR # Bug #
623192915250469
623469015251578
628551715271646
651569115373456
653479315384203
655153415393629
661696515430790
662304315434464
665422715452821
667248915463379
672209315492015
674129615504036
674600115506837
677475415524616
677475915524623
678904915532381
681085015545095
686809715580174
691945315617266


Changes incorporated in this version: 6231929 6234690 6285517 6515691 6534793 6551534 6616965 6623043 6654227 6672489 6722093 6741296 6746001 6774754 6774759 6789049 6810850 6868097 6919453

Patches accumulated and obsoleted by this patch:

Patches which conflict with this patch:

Patches required with this patch:

Obsoleted by:

Files included with this patch:

/lib/cpu/sparcv8plus+vis/libmvec_isa.so.1
/lib/cpu/sparcv9+vis/sparcv9/libmvec_isa.so.1
/lib/cpu/sparcv9+vis2/libmvec_isa.so.1
/lib/cpu/sparcv9+vis2/sparcv9/libmvec_isa.so.1
/lib/libm.so.1
/lib/libm.so.2
/lib/libm/libm_hwcap1.so.2
/lib/libm/sparcv9/libm_hwcap1.so.2
/lib/libmvec.so.1
/lib/sparcv9/libm.so.1
/lib/sparcv9/libm.so.2
/lib/sparcv9/libmvec.so.1
/usr/include/fenv.h
/usr/include/floatingpoint.h
/usr/include/iso/math_c99.h
/usr/include/math.h

Problem Description:

6231929 libm contains local symbols mdb dislikes
6234690 libm.so.2 expf(-100.0F) on US3 very slow even with -fns=yes
6285517 gcc 3.4.x/4.x c99 math intrinsics support needed
6515691 hardware-based fma and fmaf needed for SPARC boxes w/ FMAf instr.
6534793 accuracy issue with Bessel Jn function for n=6,7,8
6551534 math.h does not play well with <cstdio> w/ __EXTENSIONS__ defined
6616965 retrospective diagnostics should use dynsym/ldynsym when symbol table is not present
6623043 add vector trigpi functions to libmvec
6654227 add exception handling support for Rock instructions
6672489 SEGV when attempting to call FEX_CUSTOM exception handler
6722093 cpow(pure imaginary, real) delivered bad value
6741296 use of "exception" in math.h can cause C++ compilation failure
6746001 4882158 also affects C99 scalbln functions
6774754 tgamma(tiny) can fail to raise inexact
6774759 pow and powl can raise spurious overflow exception
6789049 sparc logbl(subnormal) delivers bogus result in C99 mode
6810850 stack trace from libm provided by fex_set_log could use some improvement
6868097 powl delivers incorrect result
6919453 Macro definitions in iso/math_c99.h header cause errors for gnu compilers


Patch Installation Instructions:
--------------------------------
 
For Solaris 8-10 releases, refer to the man pages for instructions on
using 'patchadd' and 'patchrm' scripts provided with Solaris.
Any other special or non-generic installation instructions should be
described below as special instructions.  The following example
installs a patch to a standalone machine:
 
        example# patchadd /var/spool/patch/106326-01
 
The following example removes a patch from a standalone system:
 
        example# patchrm 104945-02
 
For additional examples please see the appropriate man pages.


Special Install Instructions:
-----------------------------
none.


README -- Last modified date: Friday, May 20, 2016